From mboxrd@z Thu Jan 1 00:00:00 1970 From: mike stump To: jbuck@synopsys.com Cc: gcc@gcc.gnu.org, lerdsuwa@users.sourceforge.net Subject: Re: fix for PR 4447: is this really correct? Date: Fri, 30 Nov 2001 18:14:00 -0000 Message-ID: <200112010213.SAA17786@kankakee.wrs.com> X-SW-Source: 2001-11/msg01672.html Message-ID: <20011130181400.kDWqYUG-hBKLetIxCYegwaGguKc4Y-StR7U83_ZzEwk@z> > From: Joe Buck > To: mrs@windriver.com (mike stump) > Date: Fri, 30 Nov 2001 16:20:32 -0800 (PST) > > > But I am now not sure that this fix is quite correct, though it does > > > improve things. > > > > I think the ICE is preferable, as otherwise you have to explain that > > you have to break the ABI, which is worse. > I'd prefer to have neither the ICE nor the ABI breakage, but I'd > prefer the latter to the former. Personally, I didn't think it would be possible to maintain the abi, and I previously said as much. We can get close, but it is fairly hard. We have to decide, create more abi headaches for the future now, or not. The benefit of the headache, is, more programs can be compiled. It we are very serious about the abi, the answer must be no. If we are not as serious about it, we can put the fix into the compiler, and create the abi headache. I leave the final decision to those folks that want to make it. I just wanted to point out the consequence of the action and ensure that everone knew that we were going to purposefully create a new abi incompatibility that didn't previously exist. > Yes. Maybe it's possible to fix the bug by applying a > constant-folding operation to template arguments before the mangler > is called. They should be folded way early. > News flash: 3.1 will have a couple of minor ABI bug fixes :-( Such is life. > so it seems that we're already in a position to break the ABI, We don't just doom 3.1 to breaking, but some random future version of the compiler, with luck, it will just be 3.1.